Detections
Explore public detection logic contributed by the community across SIEM and rule languages.
1,733 detections
Filters
Last updated
All Time
Detection languages
909
387
164
127
68
Contributors
479
417
201
177
61
Categories
20,440
11,814
5,871
5,145
4,864
1,733
Platforms
822
436
329
274
206
Products / Services
498
378
118
96
79
MITRE Techniques
433
327
225
196
196
CVEs
12
12
10
10
6
IDS Classtypes
1
1
IDS Protocols
2
Detects potential AI model extraction attacks where an adversary repeatedly queries a private model's inference API. By combining high-volume request counts with high input diversity, the rule identifies behavior indicative of an adversary constructing a labeled dataset to train a substitute, replicated model.
Detects adversarial backdoor triggers (such as pixel patches, watermark-like artifacts, or malicious token sequences) injected into AI inference inputs. The rule identifies these inputs by observing a combination of suspicious trigger patterns and significant, anomalous jumps in prediction confidence directed at a specific target class, indicative of a model integrity compromise.
Detects anomalous high-volume API querying patterns against an AI model, characterized by an abnormally high number of distinct input hashes and query frequency compared to the population baseline for that model. This pattern is consistent with model extraction or stealing attacks aimed at reconstructing a proxy model.
This rule detects potentially malicious activity related to AI agent tool manipulation. It flags two primary scenarios: 1) The invocation of an AI tool shortly after an unmanaged or potentially unauthorized modification to its definition, and 2) Detection of prompt injection patterns within the output generated by an AI agent tool, which could indicate an attacker attempting to override the agent's instructions.
Detects a suspicious sequence where a user completes a device-code authentication flow followed shortly by consenting to high-risk OAuth permissions. This pattern is characteristic of consent phishing and device-code phishing attacks, where an attacker leverages these methods to gain persistent access to a user's cloud resources via malicious application tokens.
Detects potential consent phishing attacks by identifying newly registered multi-tenant Azure AD applications that receive OAuth consent from end-users within a short timeframe (10 minutes) of application registration. The rule specifically flags consents requesting high-risk delegated scopes (e.g., Mail.Read, Mail.Send, Files.ReadWrite.All, offline_access) where the application has not been verified by a trusted publisher.
Detects anomalous authentication behavior where a single session identifier is used to access three or more distinct SaaS applications within a one-hour window. This behavior often indicates the unauthorized use of a stolen session cookie or OAuth token (Pass-the-Cookie) across multiple SSO-integrated platforms.
Detects anomalous, high-volume account enumeration against the Microsoft Self-Service Password Reset (SSPR) endpoint from a single source IP. The rule identifies a pattern of mixed response codes, specifically interleaving '50034' (Account does not exist) with successful or other outcomes, which indicates the use of automated tools to validate the existence of corporate accounts via response differentiation.
The following analytic identifies the first-time execution of child processes spawned by Zoom (zoom.exe or zoom.us). It leverages Endpoint Detection and Response (EDR) data, specifically monitoring process creation events and comparing them against previously seen child processes. This activity is significant because the execution of unfamiliar child processes by Zoom could indicate malicious exploitation or misuse of the application. If confirmed malicious, this could lead to unauthorized code execution, data exfiltration, or further compromise of the endpoint.
The following analytic utilizes Windows AppLocker event logs to identify attempts to bypass application restrictions. AppLocker is a feature that allows administrators to specify which applications are permitted to run on a system. This analytic is designed to identify attempts to bypass these restrictions, which could be indicative of an attacker attempting to escalate privileges. The analytic uses EventCodes 8007, 8004, 8022, 8025, 8029, and 8040 to identify these attempts. The analytic will identify the host, full file path, and target user associated with the bypass attempt. These EventCodes are related to block events and focus on 5 attempts or more.
Detects browser extension updates that request a combination of high-risk permissions capable of account takeover (e.g., cookies, webRequest, identity) for extensions previously classified as low-risk. This pattern often indicates a supply chain compromise where a benign extension is acquired and subsequently updated with malicious capabilities.
Detects instances where a user grants OAuth application permissions to an unverified third-party application. The rule specifically monitors for high-privilege scopes such as Mail.Read, Files.ReadWrite.All, or full_access_as_app, which are commonly abused in illicit consent grant attacks to achieve persistent access to sensitive mailbox and cloud data.
Detects the installation or execution of known MALFEX malicious packages via npm or Node.js, which are used as entry points for polyglot or disguised payload delivery.
Detects anomalous use of an LLM provider API key by identifying successful API requests originating from multiple distinct countries within a one-hour time window. This activity often indicates the potential theft and concurrent unauthorized use of an API token.
Detects the registration of new tools, plugins, or connectors into an AI agent's registry, particularly from unverified publishers or outside of sanctioned deployment windows. The rule also flags cases where these newly registered tools are immediately invoked with sensitive-scope permissions (e.g., admin, secrets, write), which is consistent with indicators of AI agent supply chain compromise or unauthorized tool injection.
Detects Azure AD sign-in events indicative of 'EvilTokens' phishing-as-a-service tradecraft. The rule monitors for initial token-grant events (device-code authentication or OAuth application consent) and subsequent successful sign-ins using existing SessionIds. This rule is designed as a foundational detection component for identifying session token replay, which allows adversaries to bypass MFA and password requirements by hijacking valid session tokens.
This rule detects potential adversarial reconnaissance activities targeting AI agents via multiple communication channels (e.g., email, collaboration tools). It identifies patterns where a sender transmits messages containing common 'jailbreak' or 'instruction-override' phrases (e.g., 'ignore previous instructions', 'bypass approval') across at least three distinct channels within a 24-hour window, indicating a concerted effort to manipulate or probe agent logic.
Detects malicious instruction-override payloads, hidden HTML comments, or obscured text (e.g., zero-font size) in third-party content ingested into LLM/RAG systems, commonly used for indirect prompt injection.
Detects adversarial prompts designed to induce an Large Language Model (LLM) to disclose sensitive information, such as PII, secrets, or internal training data. This rule monitors the input prompt for extraction-based patterns and evaluates the model's response for indicators of a sensitive data leak (DLP hit).
Detects attempts to perform a Denial of Service (DoS) attack against an AI-enabled system by flooding the inference API with high volumes of requests or overly large inputs (high token counts). The rule identifies potential resource exhaustion by correlating high request rates or large input sizes with system-wide impacts, such as increased inference queue depth or excessive GPU utilization, which could impede service availability for other users.
This rule detects attempts to bypass LLM safety guardrails using adversarial prompting techniques (e.g., 'DAN' persona, hypothetical scenarios) or by identifying patterns of repeated moderation flags and refusals. It specifically looks for user prompts containing known jailbreak templates or an excessive frequency of safety-related trigger events, while excluding authorized red-team and safety evaluation accounts.
Page 6 of 83